Overview

Once telemetry is in Application Insights, use KQL to hunt failures and latency. Focus on Success, DurationMs, ResultCode, and OperationId across the standard App Insights tables.

Exam tips

Failed requests: AppRequests | where Success == false
Slow dependencies: AppDependencies | where DurationMs > …
Error rate over time: countif(Success == false) + bin(TimeGenerated, …)
Use percentile(DurationMs, 95) for P95 — better than average alone
Join exceptions to requests on OperationId

Key fields

Field	Meaning
`TimeGenerated`	When the telemetry was ingested/recorded
`Success`	Whether the operation succeeded
`DurationMs`	Latency in milliseconds
`ResultCode`	HTTP/status-style result
`Name`	Operation or dependency name
`Target`	Downstream host/resource for dependencies
`OperationId`	End-to-end correlation ID

Find failed requests

Kql

AppRequests
| where TimeGenerated > ago(1h)
| where Success == false
| project TimeGenerated, Name, ResultCode, DurationMs, OperationId
| order by TimeGenerated desc

Find slow dependencies

Dependencies are outbound calls — databases, HTTP APIs, queues. High DurationMs here often explains slow API requests upstream.

Kql

AppDependencies
| where TimeGenerated > ago(1h)
| where DurationMs > 500
| summarize Calls = count(), AvgMs = avg(DurationMs), P95 = percentile(DurationMs, 95)
    by Name, Target
| order by P95 desc

Error rate over time

Kql

AppRequests
| where TimeGenerated > ago(24h)
| summarize
    Total = count(),
    Failures = countif(Success == false)
    by bin(TimeGenerated, 1h)
| extend ErrorRate = round(100.0 * Failures / Total, 2)
| render timechart

Exceptions with request context

Kql

AppExceptions
| where TimeGenerated > ago(6h)
| join kind=inner (
    AppRequests
    | where TimeGenerated > ago(6h)
  ) on OperationId
| project TimeGenerated, ProblemId, OuterMessage, Name, ResultCode
| order by TimeGenerated desc

AI-specific explorations

Kql

// Requests tagged with model name via custom dimensions
AppRequests
| where TimeGenerated > ago(24h)
| extend model = tostring(customDimensions["gen_ai.model"])
| where isnotempty(model)
| summarize
    count(),
    avg(DurationMs),
    percentile(DurationMs, 95)
  by model, Name

Kql

// Trace messages mentioning token or embedding errors
AppTraces
| where TimeGenerated > ago(6h)
| where Message has "embedding" or Message has "token"
| project TimeGenerated, Message, SeverityLevel, OperationId

Explore logs for errors and performance